Analysing the spontaneous speech of aphasic speakers

Background: Aphasia has very serious consequences for speech production and, hence, for communication in daily life. Nevertheless, in the standard diagnostic procedures and in clinical practice, analysis of speech production in daily life is usually ignored or is restricted to the scoring of one or more variables on rating scales. Many methods for describing and analysing discourse production have been developed during the last few decades, with different aims and focusing on different linguistic levels. Aims: The aim of this paper is to present a critical review of the various methods that are currently used for the analysis of spontaneous speech (discourse) of aphasic patients. A distinction is made between semi-spontaneous speech (description of situational pictures or story telling) and real spontaneous speech (conversations or interviews based on standardised questions). Both types of connected speech can be analysed with two types of measuring instruments: rating scales and quantified linguistic variables. The advantages and disadvantages of the measuring instruments most frequently used for pragmatic, conversational, and linguistic analyses are discussed. Special attention is paid to the (ecological) validity and reliability of the methods. Main Contribution: The general outcome of the evaluation of procedures for analysing speech production and communicative abilities is that several methods do not fulfil the requirements for reliability and validity. Other methods are extremely time-consuming or give little information on the consequences of the outcomes for clinical practice. Conclusions: None of the discussed procedures is perfect, although some are better than others. Most methods do not give indications for treatment of speech production and communicative disorders in daily life and are not suitable for determining changes in the spontaneous speech of individual patients.